De Groot A, George J, Gonzalez M, Jin X, Bond K, Koup R, McNicholl J; Brown University, Providence, RI.
EpiMatrix is a matrix-based algorithm for the prediction of MHC ligands. We performed in vitro studies to test EpiMatrix selection of novel MHC ligands/CTL epitopes from HIV sequences. Binding of predicted peptides ranging from 8 to 11 amino acids in length to the surface of MHC allele specific TAP deficient cell lines was measured by FACS. Results from studies performed in three laboratories showed that 3/7 A2 peptides tested by the TB/HIV Research Lab bound (43%); 4/7 and 33/73 A11 peptides tested by the TB/HIV Research Lab and the Bond/McNicholl lab group bound, respectively (57% and 43%); 7/7 and 11/54 peptides tested by the TB/HIV Research Lab and the Jin/Koup laboratory group bound, respectively (100% and 20%); and 7/7 B27 peptides tested by the TB/HIV Research Laboratory bound (100%). The observed proportion of peptides binding was higher than expected in five out of the six sets of experiments, when compared to the median estimated binding probability (EBP) of the groups of peptides tested. Four novel HIV CTL epitopes were identified by the Jin/Koup laboratory group, and 12 novel HIV CTL epitopes were identified by the Bond/McNicholl laboratory group. Additional binding assays and CTL assays are in progress. EpiMatrix can be used to facilitate research on HIV immunopathogenesis and vaccine development.
Keywords: AEGIS, HIV, Computational Biology, Epitopes, Ligands, Major Histocompatibility Complex, Proteins, HIV Antigens, Peptides, Gene Products, gag, Gene Products, nef, Gene Products, env, HIV Infections, Algorithms, Gene Products, pol, Myosin Heavy Chains, HIV Envelope Protein gp160, in vitro, immunology, genetics, AIDS
